Lattice solid simulation of the physics of fault zones and earthquakes: The model, results and directions

The lattice solid model is a particle-based numerical model that is being developed with the aim of eventually allowing all relevant physical phenomena underlying earthquake dynamics to be realistically simulated. In its present form, the model is capable of simulating the evolution and dynamics of complex fault zones consisting of a brittle solid and/or granular gouge region. Elasticity is modeled in solid regions and friction is simulated between solid regions in contact. Solid regions can fracture and break down allowing the evolution of fault zones to be studied. Some thermal effects are simulated including thermo-mechanical and thermo-porous feedback which respectively allow studies of the effects of thermal expansion of the solid and increased pore fluid pressure when frictional heating occurs. We review the lattice solid model and discuss major results obtained so far. The model has been applied to studies of the dynamics and evolution of fault gouge zones. For inter-particle friction ranging from 0.4 to 1.0 and pressure changing by a factor of 10, the effective fault friction is found to remain approximately constant with a value of (Formula Presented) ~ 0.6. In some cases, the active slip band became highly localized after a large displacement and (Formula Presented); dropped dramatically to around 0.3. These results suggest that a self-regulation process controlling the balance between slip, rolling and fracture is responsible for the almost constant friction of 0.6 measured in the laboratory for different rock types. They also suggest it is possible for a gouge zone to self-organize to a structure in which slip is highly localized and the fault zone is anomalously weak, thus providing a comprehensive potential explanation for the heat flow paradox. Simulations show that thermal effects play a role on rupture dynamics with a relatively minor influence originating from thermal expansion and a larger effect due to the increase in pore pressure. Long simulations involving a fractured zone show accelerating Benioff strain sequences and evolving event-size statistics in the periods prior to large simulated earthquakes which are consistent with the critical point hypothesis for earthquakes. The simulation results using the present lattice solid model demonstrate the potential. of the approach to address fundamental questions regarding the physics of earthquakes from nucleation studies to studies of the dynamics of interacting fault systems. The model results suggest that earthquake forecasting or intermediate term earthquake prediction is a realistic goal and that the statistical physics analogies of the earthquake problem have relevance.